Thanks to feedback received from my earlier post on Survey Research, I purchased a few more books from the Sage Quantitative Applications in the Social Sciences series: Arlene Fink, The Survey Kit (2002) and Graham Kalton, Introduction to Survey Sampling (1983). These monographs are short and readable explanations for people interested in specific research methodologies or background theory.
While purchasing these books, I also bought another Sage monograph, Tim Hagle, Basic Math for Social Scientists: Concepts (1996), which arrived this morning. As I have gotten deeper into empirical work, I really want to fully understand the math that is driving the Stata and SPSS regression models and diagnostic tools.
Fortunately, this little book by Hagle (who is a lawyer and a political scientist) turns out to be the book that I was one day hoping to find: a quick, nuts & bolts discussion of the important math concepts that underlie probability and multivariate regression: algebra review of exponents and logarithms, limits and continuity, differential and integral calculus (which I have forgotten from college), matrix algebra, and eigenvalues and eigenvectors. It also has a handy summary of math symbols and expressions. I am very happy with this purchase.
Hagle is good (I used it myself for years in the math "prefresher" course I teach), but for those looking for a more comprehensive option, I'd suggest Jeff Gill's recent Essential Mathematics for Political and Social Research:
http://www.amazon.com/gp/product/052168403X/002-3149706-4972027?v=glance&n=283155
For people interested in doing statistics beyond regression -- including maximum likelihood-based models -- Gill's coverage is better. It's a bit more expensive than a Sage monograph, but still nowhere close to a casebook...
Posted by: Christopher Zorn | 03 April 2007 at 09:32 AM